diff --git a/DESCRIPTION b/DESCRIPTION
index e889b9c..dde3929 100644
--- a/DESCRIPTION
+++ b/DESCRIPTION
@@ -1,6 +1,6 @@
 Package: inlmisc
 Title: Miscellaneous Functions for the USGS INL Project Office
-Version: 0.4.3.9000
+Version: 0.4.4
 Authors@R: person(given=c("Jason", "C."), family="Fisher", role=c("aut", "cre"), email="jfisher@usgs.gov", comment=c(ORCID="0000-0001-9032-8912"))
 Description: A collection of functions for creating high-level graphics,
     performing raster-based analysis, processing MODFLOW-based models,
@@ -12,7 +12,7 @@ Depends:
     R (>= 3.4.0)
 Imports:
     checkmate,
-    dplyr,
+    data.table,
     GA,
     graphics,
     grDevices,
diff --git a/Makefile b/Makefile
index 9b42b1c..018316d 100644
--- a/Makefile
+++ b/Makefile
@@ -7,8 +7,8 @@ PKGSRC  := $(shell basename `pwd`)
 all: docs install check
 
 docs:
-	R -q -e 'devtools::document()';\
-	R -q -e 'devtools::clean_dll()';\
+	R -q -e 'Rd2roxygen::roxygen_and_build('\''.'\'', build=FALSE, reformat=FALSE)';\
+	R -q -e 'pkgbuild::clean_dll()';\
 
 build:
 	cd ..;\
diff --git a/NAMESPACE b/NAMESPACE
index 3db4ac9..4e9c6d7 100644
--- a/NAMESPACE
+++ b/NAMESPACE
@@ -11,6 +11,7 @@ export(AddNorthArrow)
 export(AddPoints)
 export(AddScaleBar)
 export(AddSearchButton)
+export(BuildVignettes)
 export(BumpDisconnectCells)
 export(BumpRiverStage)
 export(CreateWebMap)
@@ -44,5 +45,6 @@ export(SetPolygons)
 export(SummariseBudget)
 export(ToScientific)
 import(rgdal)
+importFrom(data.table,data.table)
 importFrom(igraph,clusters)
 useDynLib(inlmisc, .registration=TRUE, .fixes="C_")
diff --git a/NEWS.md b/NEWS.md
index 2a49013..1eb4dd8 100644
--- a/NEWS.md
+++ b/NEWS.md
@@ -1,4 +1,10 @@
-# inlmisc 0.4.3.9000
+# inlmisc 0.4.4
+
+- Add `BuildVignettes` function, used to build package vignettes.
+
+- In `SummariseBudget` function, improve memory management.
+
+- Change package imports by adding **data.table** and removing **dplyr**.
 
 - Add `SetHinge` function, used to specify a hinge location in a color palette.
 
diff --git a/R/BuildVignettes.R b/R/BuildVignettes.R
new file mode 100644
index 0000000..8840fb9
--- /dev/null
+++ b/R/BuildVignettes.R
@@ -0,0 +1,53 @@
+#' Build Package Vignettes
+#'
+#' Build package vignettes from their sources and place in the \code{/inst/doc} folder.
+#'
+#' @param pkg 'character' string.
+#'   Package path, by default the \link[=getwd]{working directory}.
+#' @param quiet 'logical' flag.
+#'   Whether to supress most output.
+#' @param clean 'logical' flag.
+#'   Whether to remove all intermediate files generated by the build.
+#' @param gs_quality 'character' string.
+#'   Quality of compacted PDF files,
+#'   see \code{\link[tools]{compactPDF}} function for details.
+#'
+#' @author J.C. Fisher, U.S. Geological Survey, Idaho Water Science Center
+#'
+#' @seealso \code{\link[tools]{buildVignettes}}
+#'
+#' @keywords utilities
+#'
+#' @export
+#'
+#' @examples
+#' \dontrun{
+#' BuildVignettes("<path/to/package>", gs_quality = "ebook")
+#' }
+#'
+
+BuildVignettes <- function(pkg=".", quiet=TRUE, clean=TRUE, gs_quality=NULL) {
+
+  checkmate::assertFileExists(file.path(pkg, "DESCRIPTION"))
+  checkmate::assertFlag(quiet)
+  checkmate::assertFlag(clean)
+  if (!is.null(gs_quality))
+    gs_quality <- match.arg(gs_quality, c("none", "printer", "ebook", "screen"))
+
+  tools::buildVignettes(dir=pkg, quiet=quiet, clean=clean, tangle=TRUE)
+
+  v <- tools::pkgVignettes(dir=pkg, output=TRUE, source=TRUE)
+  if (length(v) == 0) return(invisible(NULL))
+  out <- c(v$outputs, unique(unlist(v$sources, use.names=FALSE)))
+
+  doc <- file.path(pkg, "inst/doc")
+
+  dir.create(doc, showWarnings=!quiet, recursive=TRUE)
+  file.copy(c(v$docs, out), doc, overwrite=TRUE)
+  file.remove(out)
+
+  if (!is.null(gs_quality))
+    tools::compactPDF(paths=doc, gs_quality=gs_quality)
+
+  invisible(TRUE)
+}
diff --git a/R/ExportRasterStack.R b/R/ExportRasterStack.R
index 20fc1db..9b511b5 100644
--- a/R/ExportRasterStack.R
+++ b/R/ExportRasterStack.R
@@ -60,13 +60,11 @@ ExportRasterStack <- function(rs, path, zip="", col=NULL) {
 
   # check arguments
   stopifnot(inherits(rs, c("RasterStack", "RasterBrick")))
-  checkmate::assertDirectoryExists(path)
+  checkmate::assertString(path)
   checkmate::assertString(zip)
   if (zip != "") checkmate::assertFileExists(zip)
   checkmate::assertCharacter(col, null.ok=TRUE)
 
-  if (is.null(col)) col <- GetColors(255, stops=c(0.3, 0.9))
-
   dir.create(path, showWarnings=FALSE, recursive=TRUE)
   dir.create(path.csv <- file.path(path, "csv"), showWarnings=FALSE)
   dir.create(path.png <- file.path(path, "png"), showWarnings=FALSE)
@@ -74,6 +72,8 @@ ExportRasterStack <- function(rs, path, zip="", col=NULL) {
   dir.create(path.rda <- file.path(path, "rda"), showWarnings=FALSE)
   dir.create(path.kml <- file.path(path, "kml"), showWarnings=FALSE)
 
+  if (is.null(col)) col <- GetColors(255, stops=c(0.3, 0.9))
+
   n <- 0L
   for (i in names(rs)) {
     n <- n + 1L
diff --git a/R/GetColors.R b/R/GetColors.R
index adfd533..8320531 100644
--- a/R/GetColors.R
+++ b/R/GetColors.R
@@ -92,7 +92,12 @@
 #'   Generic Mapping Tools: Improved version released, AGU, v. 94, no. 45, p. 409--410
 #'   doi:\href{https://doi.org/10.1002/2013EO450001}{10.1002/2013EO450001}
 #'
-#' @seealso \code{\link{SetHinge}}, \code{\link[grDevices]{col2rgb}}
+#' @seealso
+#'   \code{\link{SetHinge}} function to set the hinge location in
+#'   a color palette derived from one or two color schemes.
+#'
+#'   \code{\link[grDevices]{col2rgb}} function to express palette
+#'   colors represented in the hexadecimal format as RGB triplets (R, G, B).
 #'
 #' @keywords color
 #'
diff --git a/R/ReadCodeChunks.R b/R/ReadCodeChunks.R
index 3800ab5..73af05e 100644
--- a/R/ReadCodeChunks.R
+++ b/R/ReadCodeChunks.R
@@ -25,6 +25,7 @@
 #' @export
 #'
 #' @examples
+#' \dontrun{
 #' file <- system.file("misc", "knitr-markdown.Rmd", package = "inlmisc")
 #' chunks <- ReadCodeChunks(file)
 #'
@@ -33,6 +34,7 @@
 #' chunks[["named-chunk-2"]]
 #'
 #' eval(parse(text = unlist(chunks[c("unnamed-chunk-3", "named-chunk-4")])))
+#' }
 #'
 
 ReadCodeChunks <- function(path) {
diff --git a/R/ReadModflowBinary.R b/R/ReadModflowBinary.R
index 2f68ed7..21687ed 100644
--- a/R/ReadModflowBinary.R
+++ b/R/ReadModflowBinary.R
@@ -67,9 +67,9 @@ ReadModflowBinary <- function(path, data.type=c("array", "flow"),
   endian <- match.arg(endian)
   checkmate::assertFlag(rm.totim.0)
 
-  ans <- try(.ReadBinary(path, data.type, endian, nbytes=4), silent=TRUE)
+  ans <- try(.ReadBinary(path, data.type, endian, nbytes=4L), silent=TRUE)
   if (inherits(ans, "try-error"))
-    ans <- .ReadBinary(path, data.type, endian, nbytes=8)
+    ans <- .ReadBinary(path, data.type, endian, nbytes=8L)
   if (rm.totim.0)
     ans <- ans[vapply(ans, function(i) i$totim, 0) != 0]
   ans
@@ -81,7 +81,7 @@ ReadModflowBinary <- function(path, data.type=c("array", "flow"),
   checkmate::assertFileExists(path)
   checkmate::assertString(data.type)
   checkmate::assertString(endian)
-  stopifnot(nbytes %in% c(4, 8))
+  stopifnot(nbytes %in% c(4L, 8L))
 
   con <- file(path, open="rb", encoding="bytes")
   on.exit(close(con, type="rb"))
@@ -136,33 +136,33 @@ ReadModflowBinary <- function(path, data.type=c("array", "flow"),
                     "wells")
   lst <- list()
   repeat {
-    kstp <- readBin(con, "integer", n=1, size=4, endian=endian)
+    kstp <- readBin(con, "integer", n=1L, size=4L, endian=endian)
     if (length(kstp) == 0) break
-    kper <- readBin(con, "integer", n=1, size=4, endian=endian)
+    kper <- readBin(con, "integer", n=1L, size=4L, endian=endian)
 
     if (data.type == "array") {
-      pertim <- readBin(con, "numeric", n=1, size=nbytes, endian=endian)
-      totim  <- readBin(con, "numeric", n=1, size=nbytes, endian=endian)
-      desc   <- readBin(readBin(con, "raw", n=16, size=1, endian=endian),
-                        "character", n=1, endian=endian)
+      pertim <- readBin(con, "numeric", n=1L, size=nbytes, endian=endian)
+      totim  <- readBin(con, "numeric", n=1L, size=nbytes, endian=endian)
+      desc   <- readBin(readBin(con, "raw", n=16L, size=1L, endian=endian),
+                        "character", n=1L, endian=endian)
       desc <- .TidyDescription(desc)
       if (!desc %in% valid.desc) break
-      ncol  <- readBin(con, "integer", n=1, size=4, endian=endian)
-      nrow  <- readBin(con, "integer", n=1, size=4, endian=endian)
-      layer <- readBin(con, "integer", n=1, size=4, endian=endian)
+      ncol  <- readBin(con, "integer", n=1L, size=4L, endian=endian)
+      nrow  <- readBin(con, "integer", n=1L, size=4L, endian=endian)
+      layer <- readBin(con, "integer", n=1L, size=4L, endian=endian)
       v     <- readBin(con, "numeric", n=nrow * ncol, size=nbytes, endian=endian)
       d <- matrix(v, nrow=nrow, ncol=ncol, byrow=TRUE)
       lst[[length(lst) + 1]] <- list(d=d, kstp=kstp, kper=kper, desc=desc,
                                      layer=layer, pertim=pertim, totim=totim)
 
     } else if (data.type == "flow") {
-      desc <- readBin(readBin(con, "raw", n=16, size=1, endian=endian),
-                      "character", n=1, endian=endian)
+      desc <- readBin(readBin(con, "raw", n=16L, size=1L, endian=endian),
+                      "character", n=1L, endian=endian)
       desc <- .TidyDescription(desc)
       if (!desc %in% valid.desc) break
-      ncol <- readBin(con, "integer", n=1, size=4, endian=endian)
-      nrow <- readBin(con, "integer", n=1, size=4, endian=endian)
-      nlay <- readBin(con, "integer", n=1, size=4, endian=endian)
+      ncol <- readBin(con, "integer", n=1L, size=4L, endian=endian)
+      nrow <- readBin(con, "integer", n=1L, size=4L, endian=endian)
+      nlay <- readBin(con, "integer", n=1L, size=4L, endian=endian)
 
       if (nlay > 0) {
         x <- .Read3dArray(con, nrow, ncol, nlay, nbytes, endian)
@@ -173,26 +173,25 @@ ReadModflowBinary <- function(path, data.type=c("array", "flow"),
 
       } else {  # compact form is used
         nlay <- abs(nlay)
-        itype  <- readBin(con, "integer", n=1, size=4, endian=endian)
-        delt   <- readBin(con, "numeric", n=1, size=nbytes, endian=endian)
-        pertim <- readBin(con, "numeric", n=1, size=nbytes, endian=endian)
-        totim  <- readBin(con, "numeric", n=1, size=nbytes, endian=endian)
+        itype  <- readBin(con, "integer", n=1L, size=4L, endian=endian)
+        delt   <- readBin(con, "numeric", n=1L, size=nbytes, endian=endian)
+        pertim <- readBin(con, "numeric", n=1L, size=nbytes, endian=endian)
+        totim  <- readBin(con, "numeric", n=1L, size=nbytes, endian=endian)
 
-        if (itype == 5)
-          nval <- readBin(con, "integer", n=1, size=4, endian=endian)
+        if (itype == 5L)
+          nval <- readBin(con, "integer", n=1L, size=4L, endian=endian)
         else
           nval <- 1L
         if (nval > 100) stop("more than one-hundred varaiables for each cell")
         if (nval > 1) {
-          ctmp <- readBin(readBin(con, "raw", n=16, size=1, endian=endian),
-                          "character", n=nval - 1, endian=endian)
+          ctmp <- readBin(readBin(con, "raw", n=16L, size=1L, endian=endian),
+                          "character", n=nval - 1L, endian=endian)
           ctmp <- .TidyDescription(ctmp)
         } else {
           ctmp <- NULL
         }
 
-        if (itype %in% c(0, 1)) {
-          nvalues <- ncol * nrow * nlay
+        if (itype %in% c(0L, 1L)) {
           d <- .Read3dArray(con, nrow, ncol, nlay, nbytes, endian)
           for (i in seq_along(d)) {
             lst[[length(lst) + 1]] <- list(d=d[[i]], kstp=kstp, kper=kper,
@@ -200,46 +199,51 @@ ReadModflowBinary <- function(path, data.type=c("array", "flow"),
                                            pertim=pertim, totim=totim)
           }
 
-        } else if (itype %in% c(2, 5)) {
-            nlist <- readBin(con, "integer", n=1, size=4, endian=endian)
+        } else if (itype %in% c(2L, 5L)) {
+            nlist <- readBin(con, "integer", n=1L, size=4L, endian=endian)
             if (nlist > (nrow * ncol * nlay))
               stop("large number of cells for which values will be stored")
             if (nlist > 0) {
-              d <- matrix(0, nrow=nlist, ncol=nval + 4)
+              d <- matrix(0, nrow=nlist, ncol=nval + 4L)
               colnames(d) <- make.names(c("icell", "layer", "row", "column", "flow", ctmp),
                                         unique=TRUE)
               for (i in seq_len(nlist)) {
-                d[i, 1] <- readBin(con, "integer", n=1, size=4, endian=endian)
+                d[i, 1] <- readBin(con, "integer", n=1L, size=4L, endian=endian)
                 d[i, seq_len(nval) + 4] <- readBin(con, "numeric", n=nval,
                                                    size=nbytes, endian=endian)
               }
               nrc <- nrow * ncol
-              d[, "layer"] <- as.integer((d[, "icell"] - 1) / nrc + 1)
-              d[, "row"]  <- as.integer(((d[, "icell"] - (d[, "layer"] - 1) * nrc) - 1) / ncol + 1)
-              d[, "column"] <- as.integer(d[, "icell"] - (d[, "layer"] - 1) * nrc - (d[, "row"] - 1) * ncol)
+              d[, "layer"] <- as.integer((d[, "icell"] - 1L) / nrc + 1L)
+              d[, "row"]  <- as.integer(((d[, "icell"] - (d[, "layer"] - 1L) * nrc)
+                                         - 1L) / ncol + 1L)
+              d[, "column"] <- as.integer(d[, "icell"] - (d[, "layer"] - 1L)
+                                          * nrc - (d[, "row"] - 1L) * ncol)
               lst[[length(lst) + 1]] <- list(d=d, kstp=kstp, kper=kper, desc=desc,
                                              delt=delt, pertim=pertim, totim=totim)
             }
 
-        } else if (itype == 3) {
-          layers <- readBin(con, "integer", n=nrow * ncol, size=4, endian=endian)
+        } else if (itype == 3L) {
+          layers <- readBin(con, "integer", n=nrow * ncol, size=4L, endian=endian)
           values <- readBin(con, "numeric", n=nrow * ncol, size=nbytes, endian=endian)
           for (i in sort(unique(layers))) {
             v <- values[layers == i]
             d <- matrix(v, nrow=nrow, ncol=ncol, byrow=TRUE)
             lst[[length(lst) + 1]] <- list(d=d, kstp=kstp, kper=kper, desc=desc,
-                                           layer=i, delt=delt, pertim=pertim, totim=totim)
+                                           layer=i, delt=delt, pertim=pertim,
+                                           totim=totim)
           }
 
-        } else if (itype == 4) {
+        } else if (itype == 4L) {
           v <- readBin(con, "numeric", n=nrow * ncol, size=nbytes, endian=endian)
           d <- matrix(v, nrow=nrow, ncol=ncol, byrow=TRUE)
           lst[[length(lst) + 1]] <- list(d=d, kstp=kstp, kper=kper, desc=desc,
-                                         layer=1, delt=delt, pertim=pertim, totim=totim)
+                                         layer=1L, delt=delt, pertim=pertim,
+                                         totim=totim)
           d[, ] <- 0
           for (i in seq_len(nlay)[-1]) {
             lst[[length(lst) + 1]] <- list(d=d, kstp=kstp, kper=kper, desc=desc,
-                                           layer=i, delt=delt, pertim=pertim, totim=totim)
+                                           layer=i, delt=delt, pertim=pertim,
+                                           totim=totim)
           }
 
         } else {
diff --git a/R/SetHinge.R b/R/SetHinge.R
index a2c0fc7..1b1d8a2 100644
--- a/R/SetHinge.R
+++ b/R/SetHinge.R
@@ -3,14 +3,14 @@
 #' The \emph{hinge} indicates a dramatic color change in a palette
 #' that is typically located at the midpoint of the data range.
 #' An asymmetrical data range can result in an undesired hinge location,
-#' where the location does not necessarily coincide with the break-point in the user's data.
-#' This function is used to specify a hinge location that is appropriate for your data.
+#' a location that does not necessarily coincide with the break-point in the user's data.
+#' This function can be used to specify a hinge location that is appropriate for your data.
 #'
 #' @param x 'numeric' object that can be passed to the \code{\link{range}}
 #'   function with \code{NA}'s removed.
-#'   That is, the user's data range (such as, at sea-level).
+#'   The user's data range.
 #' @param hinge 'numeric' number.
-#'   Hinge value in data units.
+#'   Hinge value (such as, at sea-level) in data units.
 #' @param scheme 'character' vector of length 1 or 2, value is recycled as necessary.
 #'   Name of color scheme(s).
 #'   The color palette is derived from one or two color schemes.
@@ -29,7 +29,7 @@
 #'   Values applied separately on either side of the hinge.
 #' @param stops 'numeric' vector of length 2.
 #'   Color stops defined by interval endpoints (between 0 and 1)
-#'   and used to select a subset of the color palette.
+#'   and used to select a subset of the color palette(s).
 #' @param allow_bias 'logical' flag.
 #'   Whether to allow bias in the color spacing.
 #'
@@ -181,7 +181,7 @@ SetHinge <- function(x, hinge, scheme="sunset", alpha=NULL, reverse=FALSE,
   s1 <- c(stp[1] + adj[1], ran - buf[1])
   s2 <- c(1 - ran + buf[2], 1 - stp[2] - adj[2])
 
-  if (s1[1] >= s1[2] | s2[1] >= s2[2])
+  if (s1[1] >= s1[2] || s2[1] >= s2[2])
     stop("problem with color stops and (or) buffer values")
 
   FUN <- function(...) {
diff --git a/R/SummariseBudget.R b/R/SummariseBudget.R
index b0c1dcb..4c80820 100644
--- a/R/SummariseBudget.R
+++ b/R/SummariseBudget.R
@@ -20,7 +20,7 @@
 #'   Data in the MODFLOW cell-by-cell budget file must be saved using the
 #'   \emph{\bold{"COMPACT BUDGET"}} output option.
 #'
-#' @return Returns a 'data.frame' with the following variables:
+#' @return Returns a 'data.table' with the following variables:
 #'   \describe{
 #'     \item{desc}{description of data type, such as "wells".}
 #'     \item{kper}{stress period}
@@ -44,6 +44,8 @@
 #'
 #' @keywords utilities
 #'
+#' @importFrom data.table data.table
+#'
 #' @export
 #'
 #' @examples
@@ -59,15 +61,15 @@ SummariseBudget <- function(budget, desc=NULL, id=NULL) {
     checkmate::assertFileExists(budget)
     budget <- ReadModflowBinary(budget, "flow")
   }
-  checkmate::assertCharacter(desc, any.missing=FALSE, min.len=1, unique=TRUE, null.ok=TRUE)
+  checkmate::assertCharacter(desc, any.missing=FALSE, min.len=1,
+                             unique=TRUE, null.ok=TRUE)
   checkmate::assertString(id, null.ok=TRUE)
 
   budget.desc <- vapply(budget, function(i) i$desc, "")
-  if (is.null(desc)) {
-    desc <- unique(budget.desc)
-  } else {
+  if (!is.null(desc)) {
     is <- desc %in% budget.desc
-    if (all(!is)) stop("data type(s) not found in budget")
+    if (all(!is))
+      stop("data type(s) not found in budget")
     if (any(!is))
       warning(sprintf("data type(s) not found in budget: %s",
                       paste(paste0("\"", desc[!is], "\""), collapse=", ")))
@@ -75,56 +77,44 @@ SummariseBudget <- function(budget, desc=NULL, id=NULL) {
   }
 
   is <- vapply(budget, function(x) !is.null(colnames(x$d)), FALSE)
-  if (all(!is)) stop("data type(s) not saved using correct form")
+  if (all(!is))
+    stop("data type(s) not saved using correct form")
   if (any(!is)) {
     x <- unique(vapply(budget[!is], function(i) i$desc, ""))
     warning(sprintf("removed data type(s): %s not saved using correct form",
-            paste(paste0("\"", x, "\""), collapse=", ")))
+                    paste(paste0("\"", x, "\""), collapse=", ")))
   }
   budget <- budget[is]
-  desc <- vapply(budget, function(i) i$desc, "")
-
-  b <- budget
-  for (i in seq_along(b)) b[[i]]$d[b[[i]]$d[, "flow"] < 0, "flow"] <- 0
-  d <- dplyr::mutate(.Summarise(b, desc, id), flow.dir="in")
-
-  b <- budget
-  for (i in seq_along(b)) b[[i]]$d[b[[i]]$d[, "flow"] > 0, "flow"] <- 0
-  d <- dplyr::bind_rows(d, dplyr::mutate(.Summarise(b, desc, id), flow.dir="out"))
-
-  d$flow.dir <- as.factor(d$flow.dir)
-  d
-}
-
 
-.Summarise <- function(b, desc, id) {
+  dt <- data.table::rbindlist(lapply(budget, function(x) {
+    d <- data.table::data.table(desc=x$desc, kper=x$kper, kstp=x$kstp, id=NA,
+                                flow=x$d[, "flow"], delt=x$delt,
+                                pertim=x$pertim, totim=x$totim)
+    if (!is.null(id) && id %in% colnames(x$d)) d$id <- x$d[, id]
+    d
+  }))
+  dt$desc <- as.factor(dt$desc)
+  dt$flow.dir <- as.character(NA)
 
-  checkmate::assertList(b)
-  checkmate::assertCharacter(desc)
-  checkmate::assertString(id, null.ok=TRUE)
+  # due to NSE notes in R CMD check
+  delt <- flow <- flow.dir <- kper <- kstp <- pertim <- totim <- NULL
 
-  d <- dplyr::bind_rows(lapply(desc, function(i) {
-    dplyr::bind_rows(lapply(b[desc == i], function(j) {
-      d <- data.frame(desc=j$desc, kper=j$kper, kstp=j$kstp, id=NA,
-                      flow=j$d[, "flow"], delt=j$delt,
-                      pertim=j$pertim, totim=j$totim, stringsAsFactors=FALSE)
-      if (!is.null(id) && id %in% colnames(j$d)) d$id <- j$d[, id]
-      d
-    }))
+  dt_summary <- data.table::rbindlist(lapply(c("in", "out"), function(i) {
+    if (i == "in")
+      dt$flow[dt$flow < 0] <- 0
+    else
+      dt$flow[dt$flow > 0] <- 0
+    dt[, list(delt        = utils::head(delt, 1),
+              pertim      = utils::head(pertim, 1),
+              totim       = utils::head(totim, 1),
+              count       = length(flow),
+              flow.sum    = sum(flow),
+              flow.mean   = mean(flow),
+              flow.median = stats::median(flow),
+              flow.sd     = stats::sd(flow),
+              flow.dir    = i),
+       by=list(desc, kper, kstp, id)]
   }))
-  d$desc <- as.factor(d$desc)
-  if ("id" %in% colnames(d))
-    grps <- dplyr::group_by_(d, "desc", "kper", "kstp", "id")
-  else
-    grps <- dplyr::group_by_(d, "desc", "kper", "kstp")
-  d <- dplyr::summarise_(grps,
-                         delt        = "delt[1]",
-                         pertim      = "pertim[1]",
-                         totim       = "totim[1]",
-                         count       = "length(flow)",
-                         flow.sum    = "sum(flow)",
-                         flow.mean   = "mean(flow)",
-                         flow.median = "stats::median(flow)",
-                         flow.sd     = "sd(flow)")
-  d
+  dt_summary$flow.dir <- as.factor(dt_summary$flow.dir)
+  dt_summary
 }
diff --git a/inst/raw/README.md b/inst/raw/README.md
index 25aac48..b721cdd 100644
--- a/inst/raw/README.md
+++ b/inst/raw/README.md
@@ -39,3 +39,9 @@ Open the *../../R/GetColors.R* file and specify the width in inches.
 ```
 
 Note that a PDF viewer is currently used to determine the layout width.
+
+## Permissions
+
+* Thomas Dewez (SRON) granted permission to add his color schemes on October 17, 2018.
+
+* Paul Tol (SRON) granted permission to add his color schemes on October 17, 2018.
diff --git a/inst/raw/internal-datasets.R b/inst/raw/internal-datasets.R
index 9a79714..4987b31 100644
--- a/inst/raw/internal-datasets.R
+++ b/inst/raw/internal-datasets.R
@@ -1,30 +1,13 @@
 MakeSysdata <- function() {
 
   options(stringsAsFactors=FALSE)
+  dir.create(file.path(getwd(), "cpt"), showWarnings=FALSE)
 
   cite <- c("Dewez"  = "Thomas Dewez (2004) grants permission to distribute with attribution.",
             "Tol"    = "Paul Tol (2018) grants permission to distribute with attribution.",
             "Wessel" = "Wessel and others (2013) released under an open license.")
 
-  exclude <- scan(what="character", quiet=TRUE, text="
-                  cool
-                  grayC
-                  polar
-                  red2green
-                  seis
-                  ")
-  diverge <- scan(what="character", quiet=TRUE, text="
-                  berlin
-                  broc
-                  cork
-                  lisbon
-                  oleron
-                  roma
-                  split
-                  tofino
-                  vik
-                  ")
-  schemes <- .GetGMTCpt(cite["Wessel"], exclude, diverge)
+  schemes <- .GetCptGmt(cite["Wessel"])
 
   schemes[["DEM screen"]] <- list(
     data = read.csv(strip.white=TRUE, text="
@@ -501,7 +484,8 @@ MakeSysdata <- function() {
 }
 
 
-.GetGMTCpt <- function(cite, exclude=NULL, diverge=NULL) {
+.GetCptGmt <- function(cite) {
+  checkmate::assertString(cite)
 
   # code adapted from stackoverflow answer by lukeA, accessed October 27, 2018
   # at https://stackoverflow.com/questions/25485216
@@ -514,24 +498,22 @@ MakeSysdata <- function() {
   httr::stop_for_status(info)
   tree <- unlist(lapply(httr::content(info)$tree, "[", "path"), use.names=FALSE)
   path <- grep("share/cpt/", tree, value=TRUE, fixed=TRUE)
-
   host <- "raw.githubusercontent.com"
   file <- sprintf("https://%s/%s/%s/master/%s", host, owner, repo, path)
 
   nm <- tools::file_path_sans_ext(basename(file))
+  exclude <- c("cool", "grayC", "polar", "red2green", "seis")
   file <- file[!nm %in% exclude]
 
-  destdir <- file.path(getwd(), "cpt")
-  dir.create(destdir, showWarnings=FALSE)
-
-  destfile <- file.path(destdir, basename(file))
+  destfile <- file.path(getwd(), "cpt", basename(file))
   for (i in seq_along(file)) {
     utils::download.file(file[i], destfile[i], quiet=TRUE)
   }
 
   nm <- tools::file_path_sans_ext(basename(file))
   type <- rep("Sequential", length(nm))
-  type[nm %in% diverge] <- "Diverging"
+  div <- c("berlin", "broc", "cork", "lisbon", "oleron", "roma", "split", "tofino", "vik")
+  type[nm %in% div] <- "Diverging"
 
   cpt <- lapply(seq_along(destfile), function(i) {
     .ReadCpt(destfile[i], cite=cite, type=type[i])
diff --git a/man/BuildVignettes.Rd b/man/BuildVignettes.Rd
new file mode 100644
index 0000000..e3105aa
--- /dev/null
+++ b/man/BuildVignettes.Rd
@@ -0,0 +1,39 @@
+% Generated by roxygen2: do not edit by hand
+% Please edit documentation in R/BuildVignettes.R
+\name{BuildVignettes}
+\alias{BuildVignettes}
+\title{Build Package Vignettes}
+\usage{
+BuildVignettes(pkg = ".", quiet = TRUE, clean = TRUE,
+  gs_quality = NULL)
+}
+\arguments{
+\item{pkg}{'character' string.
+Package path, by default the \link[=getwd]{working directory}.}
+
+\item{quiet}{'logical' flag.
+Whether to supress most output.}
+
+\item{clean}{'logical' flag.
+Whether to remove all intermediate files generated by the build.}
+
+\item{gs_quality}{'character' string.
+Quality of compacted PDF files,
+see \code{\link[tools]{compactPDF}} function for details.}
+}
+\description{
+Build package vignettes from their sources and place in the \code{/inst/doc} folder.
+}
+\examples{
+\dontrun{
+BuildVignettes("<path/to/package>", gs_quality = "ebook")
+}
+
+}
+\seealso{
+\code{\link[tools]{buildVignettes}}
+}
+\author{
+J.C. Fisher, U.S. Geological Survey, Idaho Water Science Center
+}
+\keyword{utilities}
diff --git a/man/GetColors.Rd b/man/GetColors.Rd
index 609f631..6d6a2e2 100644
--- a/man/GetColors.Rd
+++ b/man/GetColors.Rd
@@ -206,7 +206,11 @@ Dewez, Thomas, 2004, Variations on a DEM palette, accessed October 15, 2018 at
   doi:\href{https://doi.org/10.1002/2013EO450001}{10.1002/2013EO450001}
 }
 \seealso{
-\code{\link{SetHinge}}, \code{\link[grDevices]{col2rgb}}
+\code{\link{SetHinge}} function to set the hinge location in
+  a color palette derived from one or two color schemes.
+
+  \code{\link[grDevices]{col2rgb}} function to express palette
+  colors represented in the hexadecimal format as RGB triplets (R, G, B).
 }
 \author{
 J.C. Fisher, U.S. Geological Survey, Idaho Water Science Center
diff --git a/man/ReadCodeChunks.Rd b/man/ReadCodeChunks.Rd
index 998e614..464b787 100644
--- a/man/ReadCodeChunks.Rd
+++ b/man/ReadCodeChunks.Rd
@@ -27,6 +27,7 @@ If the source document is \file{.Rnw} or \file{.Rmd} the \code{purl} function is
   names like \code{unnamed-chunk-i} where \code{i} is the chunk number.
 }
 \examples{
+\dontrun{
 file <- system.file("misc", "knitr-markdown.Rmd", package = "inlmisc")
 chunks <- ReadCodeChunks(file)
 
@@ -35,6 +36,7 @@ names(chunks)
 chunks[["named-chunk-2"]]
 
 eval(parse(text = unlist(chunks[c("unnamed-chunk-3", "named-chunk-4")])))
+}
 
 }
 \seealso{
diff --git a/man/SetHinge.Rd b/man/SetHinge.Rd
index b706af6..d02dc66 100644
--- a/man/SetHinge.Rd
+++ b/man/SetHinge.Rd
@@ -10,10 +10,10 @@ SetHinge(x, hinge, scheme = "sunset", alpha = NULL, reverse = FALSE,
 \arguments{
 \item{x}{'numeric' object that can be passed to the \code{\link{range}}
 function with \code{NA}'s removed.
-That is, the user's data range (such as, at sea-level).}
+The user's data range.}
 
 \item{hinge}{'numeric' number.
-Hinge value in data units.}
+Hinge value (such as, at sea-level) in data units.}
 
 \item{scheme}{'character' vector of length 1 or 2, value is recycled as necessary.
 Name of color scheme(s).
@@ -37,7 +37,7 @@ Values applied separately on either side of the hinge.}
 
 \item{stops}{'numeric' vector of length 2.
 Color stops defined by interval endpoints (between 0 and 1)
-and used to select a subset of the color palette.}
+and used to select a subset of the color palette(s).}
 
 \item{allow_bias}{'logical' flag.
 Whether to allow bias in the color spacing.}
@@ -50,8 +50,8 @@ Returns a 'function' that takes an 'integer' argument
 The \emph{hinge} indicates a dramatic color change in a palette
 that is typically located at the midpoint of the data range.
 An asymmetrical data range can result in an undesired hinge location,
-where the location does not necessarily coincide with the break-point in the user's data.
-This function is used to specify a hinge location that is appropriate for your data.
+a location that does not necessarily coincide with the break-point in the user's data.
+This function can be used to specify a hinge location that is appropriate for your data.
 }
 \examples{
 Pal <- SetHinge(x = c(-3, 7), hinge = 0)
diff --git a/man/SummariseBudget.Rd b/man/SummariseBudget.Rd
index 84c9c82..a3f4e74 100644
--- a/man/SummariseBudget.Rd
+++ b/man/SummariseBudget.Rd
@@ -20,7 +20,7 @@ Name of auxiliary variable, a variable of additional values associated with each
 saved using the \emph{\bold{"AUXILIARY"}} output option.}
 }
 \value{
-Returns a 'data.frame' with the following variables:
+Returns a 'data.table' with the following variables:
   \describe{
     \item{desc}{description of data type, such as "wells".}
     \item{kper}{stress period}
diff --git a/man/figures/table.pdf b/man/figures/table.pdf
index adf4832..104f362 100644
Binary files a/man/figures/table.pdf and b/man/figures/table.pdf differ